Second cousin once removed and twice removed, explained with a chart
A second cousin once removed is your parent's second cousin or your second cousin's child; twice removed adds one more generation. A chart and DNA figures.

A second cousin once removed is either your parent's second cousin or your second cousin's child: one generation apart from you, descended from a couple who sit three generations above one of you and four above the other. A second cousin twice removed is two generations apart: your grandparent's second cousin, or your second cousin's grandchild. Each removal halves the expected share of DNA, from 3.125 percent for second cousins to about 1.56 percent once removed and about 0.78 percent twice removed.
Those are the definitions. The rest of this piece shows how to work the names out for any branch of a tree, where the two removals sit on a cousin chart, and why DNA results for these relatives so often wander away from the textbook numbers.
Name any removed cousin in four steps
- Find the nearest ancestors the two people both descend from. Usually that is a married couple.
- Count down each line separately. A child of the couple is 1, a grandchild 2, a great-grandchild 3.
- The cousin number is the smaller count minus one.
- The number of times removed is the difference between the two counts.
Here is a hypothetical family to practice on. Walter and Ida have two children, Grace and Henry. Their descendants run down five generations:
| Generations below Walter and Ida | Grace's line | Henry's line |
|---|---|---|
| 1 | Grace | Henry |
| 2 | Ivan, Grace's son | June, Henry's daughter |
| 3 | Kate, Ivan's daughter | Leo, June's son |
| 4 | Max, Kate's son | Owen, Leo's son |
| 5 | Nora, Max's daughter |
Kate and Leo are both 3 down: 3 - 1 = 2, second cousins, nothing removed. Now take Leo and Max. The counts are 3 and 4. The smaller is 3, so they are second cousins, and the difference is 1, so once removed. Leo and Nora: counts 3 and 5, second cousins, twice removed.
The steps work from either end, and the name is the same both ways. Max is Leo's second cousin once removed, and Leo is Max's. In kin-type letters (F father, M mother, B brother, Z sister, S son, D daughter, read outward from the first person), Leo is Max's MFMBDS: mother's father's mother's brother's daughter's son. Six letters for one relative is a fair argument for the English names, and our notation guide explains the letters if you want to use them anyway.
The second-cousin corner of the cousin chart
A cousin chart is a grid of those two counts. Pick the row for your count and the column for theirs. This corner of it starts where second cousins begin, at 3 generations down:
| You are … below the shared couple | They are 3 | They are 4 | They are 5 | They are 6 |
|---|---|---|---|---|
| 3 | second cousin | second cousin once removed | second cousin twice removed | second cousin three times removed |
| 4 | second cousin once removed | third cousin | third cousin once removed | third cousin twice removed |
| 5 | second cousin twice removed | third cousin once removed | fourth cousin | fourth cousin once removed |
Two things stand out. In the full grid every removed name appears twice, mirrored across the diagonal, because each "removed" relationship has an older side and a younger side. And the second-cousin names run along the edge of the grid: once both counts reach 4, the people are at least third cousins, however many removals follow. The cousin calculator fills in the rest of the grid, up to twelve generations each way.
Older or younger: two faces of each name
| Name | When they are in an older generation | When they are in a younger generation |
|---|---|---|
| second cousin once removed | your parent's second cousin; you share their great-grandparents, who are your great-great-grandparents | your second cousin's child; you share your great-grandparents, who are their great-great-grandparents |
| second cousin twice removed | your grandparent's second cousin; their great-grandparents are your 3rd great-grandparents | your second cousin's grandchild; your great-grandparents are their 3rd great-grandparents |
The older side is where people usually meet these relatives on paper: in a grandparent's address book, an obituary, a reunion photograph. The younger side is where they meet them in person. Either way the shared ancestors are the same couple, and only the number of "greats" changes depending on whose side you count from.
Relatedness, shared DNA and degrees
The cousin calculator turns the two counts into numbers. Wright's coefficient of relationship, r, gives each path through a shared ancestor a value of one half per parent and child link, and a couple supplies two paths. For Leo and Max the path from Max up to Walter and down to Leo has 4 + 3 = 7 links:
r = 2 × (1/2)^7 = 2/128 = 1/64
For Leo and Nora there are 8 links, so r = 2 × (1/2)^8 = 1/128. Expected shared centimorgans are 2r times the calculator's 3,500 cM per genome copy: 2 × 1/64 × 3,500 = 109.375 and 2 × 1/128 × 3,500 = 54.6875.
| Relationship | Counts | r | Expected shared DNA | Expected cM | Civil degree | Traditional canon degree |
|---|---|---|---|---|---|---|
| second cousin | 3 and 3 | 1/32 | 3.125% | about 219 | 6 | 3 |
| second cousin once removed | 3 and 4 | 1/64 | 1.5625% | about 109 | 7 | 4 |
| second cousin twice removed | 3 and 5 | 1/128 | 0.78125% | about 55 | 8 | 5 |
| third cousin | 4 and 4 | 1/128 | 0.78125% | about 55 | 8 | 4 |
| second cousin three times removed | 3 and 6 | 1/256 | 0.390625% | about 27 | 9 | 6 |
Notice the third cousin. A second cousin twice removed and a third cousin have exactly the same expectation, 1/128, because both are linked by eight parent and child links through a couple. The same coincidence ties a second cousin once removed to a half second cousin, a half first cousin twice removed and a first cousin three times removed: four different positions in a tree, all at r = 1/64.
What testers actually share
The Shared cM Project, run by the genetic genealogist Blaine Bettinger, gathers shared totals from people who know their relationship. Its version 4.0 report (March 2020) holds more submissions for the second cousin once removed than for any other relationship: 5,281 pairs, sharing between 14 and 353 cM, with an average of 122. For second cousins twice removed it lists 1,404 pairs, between 0 and 244 cM, average 71. Third cousins, for comparison: 4,775 pairs, 0 to 234 cM, average 73.
Two features of these figures need care. First, the minimum for the twice-removed relationship is zero. Some pairs of documented second cousins twice removed share no DNA that a testing company can detect, and that disproves nothing about the tree.
Second, the project's averages for relationships with a minimum of zero were calculated only over pairs who shared some DNA. That pushes them above the true average. Ten hypothetical pairs show how. Suppose three share nothing detectable and the other seven share 40, 45, 55, 60, 70, 85 and 100 cM, a total of 455. Averaged over all ten pairs, that is 455 / 10 = 45.5 cM. Averaged over the seven who matched, it is 455 / 7 = 65 cM. So the project's 71 and the calculator's 55 are not in conflict; they answer different questions. The report adds other reasons observed totals drift from theory: endogamy and pedigree collapse raise sharing, and each company's matching threshold may affect totals at distant relationships.
The scatter itself also grows, in relative terms, with every generation. W. G. Hill and B. S. Weir (2011) showed that although the variance of the actual share falls as relatives get more distant, its coefficient of variation rises, and the distribution becomes more skewed. Their conclusion, in plain words: the further out you go, the harder it is to tell one pedigree relationship from another by the share of genome alone. At the level of second cousins once and twice removed, a DNA total narrows the possibilities and the documents have to do the rest.
An older name for the same relative
English has not always said "removed". Church lawyers once described unequal lines by naming both of them. Richard Burtsell's 1908 article on consanguinity in the Catholic Encyclopedia, describing the marriage rules of some Eastern churches, glosses the seventh civil degree as "second cousins touching third". One line is three generations long, the other four: in modern genealogy, a second cousin once removed. The older phrase has one merit: it says that one line has the length of a second-cousin line and the other the length of a third-cousin line, which is exactly the half-step that "once removed" describes.
A related slip is common enough to name. Your second cousin's child is not your third cousin; your child and your second cousin's child are third cousins to each other. If you want to see how the first of these steps works, our piece on first cousin once removed vs second cousin walks through the same logic one generation up.
Check yourself
What is your grandmother's second cousin's grandchild to you?
Count from the couple your grandmother shares with her second cousin. Your grandmother is 3 generations below them, so you are 5. Her second cousin is also 3 below, so that cousin's grandchild is 5. Both counts are 5: 5 - 1 = 4, fourth cousins, nothing removed. Put 5 and 5 into the calculator and it agrees, with r = 2 × (1/2)^10 = 1/512, an expected 0.1953125 percent of DNA in common, about 14 cM. Fourth cousins share a couple five generations back, your 3rd great-grandparents, and the Shared cM Project's range for fourth cousins starts at zero.
Sources and further reading
- Bettinger, Blaine T. 2020. The Shared cM Project, Version 4.0 (March 2020). The Genetic Genealogist. https://thegeneticgenealogist.com/wp-content/uploads/2020/03/Shared-cM-Project-Version-4.pdf
- Perl, Jonny, and Blaine T. Bettinger. 2020. Shared cM Project tool, version 4 data. DNA Painter. https://dnapainter.com/tools/sharedcmv4
- Hill, W. G., and B. S. Weir. 2011. "Variation in actual relationship as a consequence of Mendelian sampling and linkage." Genetics Research 93 (1): 47-64. https://doi.org/10.1017/S0016672310000480
- Wright, Sewall. 1922. "Coefficients of inbreeding and relationship." The American Naturalist 56 (645): 330-338. https://doi.org/10.1086/279872
- Burtsell, Richard. 1908. "Consanguinity (in Canon Law)." In The Catholic Encyclopedia, vol. 4. New York: Robert Appleton Company. https://www.newadvent.org/cathen/04264a.htm